Do not spend the first trial order proving only that the supplier can make its best sample.
Use it to test the handover your factory will repeat.
Select a representative design and clear objectives
Choose a representative design. The easiest core in the catalogue may reveal little about the joint arrangement, packet count or winding interface that matters in your real programme. The most difficult design can also be a poor first trial if it changes too many unqualified variables at once.
Agree the learning objectives before production starts. One objective is dimensional: can the delivered geometry be measured from the same datums your assembly team uses? Another is magnetic: can the supplier reproduce the agreed excitation and report loss and current on a clear boundary?
A third objective is often missed: can your people use the delivery without interpreting undocumented choices? Packet identity, assembly order, removable-yoke instructions and clamp interfaces should be understandable to the receiving team, not only to the operator who built the sample.
Retain first-pass evidence and repeat the measurement
Retain the first readings as well as the released result. A core that required several unrecorded adjustments is not equivalent to one that passed through the intended route. Rework may be acceptable during development, but it must be visible if the trial is meant to qualify production.
Include a measurement repeat. Repeat the test with the same setup before attributing any later difference to handling or reassembly. Where the actual production route includes reopening the magnetic circuit, plan a controlled reassembly check with agreed responsibilities and limits.
The trial should also test technical communication. Raise a real drawing ambiguity during review and examine how it is resolved. Does the supplier ask for the functional requirement, document the decision and incorporate it into the released revision? Fast replies without a controlled decision are not the same as engineering support.
Turn trial findings into the next-order baseline
A single order cannot establish long-term batch consistency. It can establish the baseline, the traceability chain and the controls that later orders must reproduce. Subsequent production should confirm that the qualified route survives different material lots and normal scheduling conditions.
Close the trial with explicit decisions: approved scope, unresolved risks, required changes and the evidence needed on the next delivery. Avoid a vague “sample accepted” that everybody interprets differently.
The first order earns its value when the second order requires fewer assumptions, not merely when the first certificate looks impressive.
Practical takeaway
A trial should qualify a repeatable handover rather than a specially prepared sample.
Chenfan Electric manufactures custom transformer cores according to customer drawings. Where applicable to the specified material and core design, agreed process controls include burr height below 0.02 mm and stacking factor above 97%. Measurement methods and acceptance conditions are defined for each order.